Causes and Timing of Future Biosphere Extinctions

Home , Extinction

Biogeosciences, 3, 85–92, 2006 www.biogeosciences.net/3/85/2006/ Biogeosciences © Author(s) 2006. This work is licensed under a Creative Commons License.

Causes and timing of future biosphere extinctions

S. Franck, C. Bounama, and W. von Bloh Potsdam Institute for Climate Impact Research, Potsdam, Germany Received: 28 September 2005 – Published in Biogeosciences Discuss.: 7 November 2005 Revised: 25 January 2006 – Accepted: 26 January 2006 – Published: 10 March 2006

Abstract. We present a minimal model for the global carbon boiling-point of water. Within the framework of Earth sys- cycle of the Earth containing the reservoirs mantle, ocean tem science (Franck et al., 2000, 2002) our planet is de- floor, continental crust, biosphere, and the kerogen, as well scribed as a system of certain interacting components (man- as the combined ocean and atmosphere reservoir. The model tle, oceanic crust, continental lithosphere, atmosphere, hy- is specified by introducing three different types of biosphere: drosphere, and biosphere) that develops under increasing procaryotes, eucaryotes, and complex multicellular life. Dur- external forcing (increasing insolation) and changing inter- ing the entire existence of the biosphere procaryotes are al- nal forcing (decreasing spreading rate, growing continen- ways present. 2 Gyr ago eucaryotic life first appears. The tal area). Within certain limits the Earth system is able to emergence of complex multicellular life is connected with an self-regulate against changing external and internal forcing. explosive increase in biomass and a strong decrease in Cam- The life span of the biosphere is related to these limits of brian global surface temperature at about 0.54 Gyr ago. In self-regulation. Lovelock and Whitfield (1982) published the long-term future the three types of biosphere will die out the first estimations of the biosphere’s life span. Accord- in reverse sequence of their appearance. We show that there ing to their semi-quantitative model, photosynthesis ceases is no evidence for an implosion-like extinction in contrast to already in about 100 Myr because the atmospheric carbon the Cambrian explosion. In dependence of their temperature dioxide content falls below the minimum value for C3-plants tolerance complex multicellular life and eucaryotes become (150 ppm). The first quantitative model for the long-term fu- extinct in about 0.8–1.2 Gyr and 1.3–1.5 Gyr, respectively. ture of the biosphere was proposed by Caldeira and Kast- The ultimate life span of the biosphere is defined by the ex- ing (1992). With the help of a more sensitive climate model tinction of procaryotes in about 1.6 Gyr. and under the assumption of a minimum atmospheric CO2 value of 10 ppm for C4-plants, they calculated that the biosphere’s life span extends up to about 800 Myr. Franck et al. (2000) developed an Earth system model that takes into 1 Introduction account quantitatively the internal forcing by geodynamics. The general basis of this paper is the long-term evolution of This effect results in a reduction of the biosphere life span the global carbon cycle from the Archaean up to about 2 Gyr from 800 Myr to 600 Myr. The biotic enhancement of weath- into the future and its consequences for the Earth’s climate ering and its influence on the life span was investigated by and the biosphere. In particular, we investigate the influ- Lenton and von Bloh (2001). According to their results the ence of geosphere-biosphere interactions on the life span of current biosphere should remain resilient to carbon cycle per- the biosphere. The problem of the long-term existence of turbation or mass extinction events for at least 800 Myr and the biosphere was first discussed by astrophysicists. They may survive for up to 1.2 Gyr. The question of the life span analysed increase of insolation during Sun’s evolution on the of the biosphere is also connected to the question of the fate main sequence. Already in the 1960s, Unsold¨ (1967) pre- of the Earth’s ocean. Bounama et al. (2001) have shown that dicted the ultimate end of terrestrial life in about 3.5 Gyr liquid water will be always available in the surface reservoirs when solar luminosity will be about 40% higher than now as a result of internal processes. The extinction of the bio- and temperatures at the Earth’s surface will be above the sphere will not be caused by the catastrophic loss of water but by other limiting factors caused by the external forcing Correspondence to: C. Bounama of increasing solar luminosity. ([email protected])

Published by Copernicus GmbH on behalf of the European Geosciences Union. 86 S. Franck et al.: Causes and timing of future biosphere extinctions

dC All these estimations of the biosphere life span deal with f = + − −1 Fprec Co+a,Cc Fhyd τf Cf (4) a rather simple unique biosphere existing within a certain dt temperature tolerance window and above a certain minimum value of atmospheric CO2 content. A natural extension to dCbio,1 −1 = 5 (Co+a) − τ C a more specific biosphere is to introduce three types of bio- dt 1 bio,1 bio,1 sphere (procaryotes, eucaryotes, complex multicellular life) . with different temperature tolerance windows and different . (5.1...5.n) biotic enhancement of weathering. According to Ward and dCbio,n −1 = 5n (Co+a) − τ C Brownlee (2002) the long-term future of the biosphere is in dt bio,n bio,n some sense a mirror image of the history: the different biosphere types will become extinct in reverse sequence of their n appearance. dCker X −1 −1 = γ τ C ,i − τ C (6) In the present paper we apply our general model for the dt bio,i bio ker ker i=1 long-term co-evolution of the geosphere and the biosphere (Franck et al., 2002) with three different biosphere pools The variable t is the time, τf the residence time of carbon in (procaryotes, simple eucaryotes, and complex multicellular the seafloor, A the accretion ratio of carbon, R the regassing life) to investigate the long-term evolution of the biosphere. ratio, SA the areal spreading rate, fc the degassing fraction Our model was previously used to investigate the Cam- of carbon, dm the melt generation depth, Vm the mantle vol- brian explosion as triggered by geosphere-biosphere feed- ume, Fweath the weathering rate, Fprec the rate of carbonate backs (von Bloh et al., 2003). We found that the Cam- precipitation, Fhyd the hydrothermal flux, γ the fraction of brian explosion was mainly driven by extrinsic environmen- dead biomass transferred to the kerogen, τbio,i the residence tal causes and so rapid because of a positive feedback be- time of carbon in the type i biosphere, 5i the total produc- tween the spread of biosphere, increased silicate weathering, tivity of the type i biosphere, and τker is the residence time and a consequent cooling of the climate. of carbon in the kerogen. The accretion ratio, A, is defined The main questions to be answered in the following are: as the fraction of seafloor carbonates accreted to the conti- What are the life spans of the three different types of bio- nents to the total seafloor carbonates. The regassing ratio, sphere and what are the reasons for their extinction? R, is defined as the fraction of seafloor carbonates regassed into the mantle to the total subducting carbonates. SA and dm are calculated from a parameterized thermal evolution 2 Model description model of whole mantle convection including the water exchange between mantle and surface reservoirs (Franck and The global carbon cycle model of Franck et al. (2002) de- Bounama, 1995). The box model including the pertinent scribes the evolution of the mass of carbon in the mantle, fluxes is sketched in Fig. 1. Cm, in the combined reservoir consisting of ocean and atmosphere, Co+a, in the continental crust, Cc, in the ocean 2.1 Climate modelling crust and floor, Cf , in the kerogen, Cker, and in the different biospheres, Cbio,i(i=1, . . . , n), where n is the number In order to calculate the surface temperature, Ts, we need of the distinct parameterized biosphere types. The equations a climate model, which links the temperature to the given for the efficiency of carbon transport between reservoirs take partial pressure of atmospheric CO2 and the solar constant, into account mantle de- and regassing, carbonate precipita- S. We apply the grey atmosphere model of Lenton (2000). tion, carbonate accretion, evolution of continental biomass, The temperature is determined using energy balance between the storage of dead organic matter, and weathering processes. the incoming and outgoing radiation: dCm −1 4 (1 − a) S 3 = τ (1 − A) RCf − SA fc dm Cm/Vm (1) σ T = 1 + τ , (7) dt f s 4 4 where σ is the Stefan-Boltzmann constant, a is the aver- dC o+a = −1 − − + + age planetary albedo, and τ is the vertical opacity of the τf (1 A)(1 R) Cf SA fc dm Cm/Vm dt greenhouse atmosphere. The opacities of the two greenhouse n + + − X −1 + −1 − gases, CO2 and H2O, are assumed to be independent from Fweath (Co+a,Cc) (1 γ ) τbio,iCbio,i τker Cker i=1 each other: n X τ = τ pCO + τ pH O . (8) − 5i (Co+a) − Fprec (Co+a,Cc) − Fhyd (2) 2 2 i=1 The opacity of CO2 is assumed to be a function of its mix- ing ratio. It is derived from the results of varying CO2 in dCc −1 = τ AC − F (C + ,C ) (3) dt f f weath o a c a radiative-convective climate model (Kasting et al., 1993).

Biogeosciences, 3, 85–92, 2006 www.biogeosciences.net/3/85/2006/ S. Franck et al.: Causes and timing of future biosphere extinctions 87

Fig. 1. Diagram illustrating the basic mechanisms and interactions of the global carbon cycle. The ﬂuxes from and to the different pools are indicated by arrows.

c s The partial pressure of H2O, pH2O, can be expressed as a the carbonate and silicate weathering rate, Fweath and Fweath, function of temperature and relative humidity, H , using the respectively: Clausius-Clapeyron equation. Here we use a wet greenhouse c c model with H =1 (full saturation). The partial pressure of Fweath = β · fweath, (9) CO2, pCO2 , can be calculated from the Co+a reservoir under the condition of equal partial pressures of CO2 at the in- s s Fweath = β · fweath, (10) terface between atmosphere and ocean (Franck et al., 2002). The solar constant evolves according to Caldeira and Kasting c,s where f denote the original carbonate and silicate (1992). weath weathering rate without additional biotic enhancement. The Our climate model depends only on the brightening Sun prefactor β reflects the biotic enhancement of weathering by and the CO /H O greenhouse effect. Therefore, we ne- 2 2 the biosphere types, i: glect so-called anti-greenhouse effects that potentially cool the planet: sulfuric acid aerosols, hydrocarbon stratospheric n Y 1 1 5i hazes, ice-albedo feedbacks, and clouds. β = + 1 − . (11) β β 5∗ i=1 i i i 2.2 Weathering rates The factor βi denotes the specific biotic amplification of ∗ There are two main types of weathering processes: silicate weathering, 5i the specific biological productivity, and 5i weathering and carbonate weathering. Both types are en- the respective present-day value of biosphere type i. In hanced by the biosphere. First, there is an increase of soil our study as a first approximation we considered a biotic CO2 partial pressure due to respiration of soil organisms and enhancement of weathering only by complex multicellular due to the respiration from the roots of vascular plants. Fur- life (β1=β2=1, β3>1). According to Schwartzman (1999, thermore, there is an additional direct dependence of weath- Figs. 8–3) complex multicellular life contributes about 7 ering on biological productivity (due to the secretion of or- times more to the biotic enhancement of weathering than ganic acids, chelating agents, etc.) by a factor β mediating primitive life. www.biogeosciences.net/3/85/2006/ Biogeosciences, 3, 85–92, 2006 88 S. Franck et al.: Causes and timing of future biosphere extinctions

Table 1. Model constants for the three different biosphere types: (1) procaryotes, (2) eucaryotes, (3) complex multicellular life. Case 1 denotes the restrictive temperature tolerance windows given by von Bloh et al. (2003), while case 2 denotes the values given by Schwartzman (1999).

Case 1 Case 2 Biosphere type i=1 i=2 i=3 i=1 i=2 i=3 ◦ Tmin ( C) 2 5 0 0 0 0 ◦ Tmax ( C) 100 45 30 100 60 45 5max (Gt/yr) 20 20 20 20 20 20 −6 Pmin (10 bar) 10 10 10 10 10 10 −6 P1/2 (10 bar) 210.8 210.8 210.8 210.8 210.8 210.8 τbio (yr) 12.5 12.5 12.5 12.5 12.5 12.5 β 1 1 3.6 1 1 17.5

2.3 Biological productivity

In our model the biological productivity is based on photo- synthetic activity and depends on the mean global surface temperature, Ts, and on the CO2 partial pressure of the atmosphere, pCO2 : = 5i 5max,i fTs,i (Ts) fCO2,i pCO2 , (12) where 5max,i is the maximum productivity of biosphere type i. The values for τbio,iand 5max,i have been adjusted to yield the present day distribution of biomass among the three different pools. The function describing the temperature depen- Fig. 2. Case 1, (a) Evolution of global surface temperature (solid dence, fT s,i , is parameterized by a parabola: green line). The green dashed line denotes a second possible evolu- tionary path triggered by a temperature perturbation in the Neopro- Ts − Tmin,i Tmax,i − Ts terozoic era. The coloured area indicates the evolution of the nor- f (T ) = 1 − (13) Ts ,i s 2 malized continental area according to Condie (1990). (b) Evolution 4 T ,i − T ,i max min of the cumulative biosphere pools for procaryotes (red), eucaryotes (green), and complex multicellular life (brown). and the function for the pCO2 dependence is a Michaelis- Menten hyperbola: p − p f p = CO2 min,i . (14) 3 Results and discussion CO2,i CO2 + − p1/2,i pCO2 pmin,i pmin,idenotes the minimum CO2 atmospheric partial The global carbon cycle model given in Eqs. (1–6) has been pressure allowing photosynthesis of biosphere type i. solved numerically for three biosphere types: procaryotes, p1/2,i+pmin,i is the pressure resulting a productivity half its simple eucaryotes (protista), and complex multicellular life. maximum value. The interval [Tmin,i...Tmax,i] denotes the The model runs have been performed for two cases: temper- temperature tolerance window. It must be emphasized that ature tolerances (case 1) given by von Bloh et al. (2003) and this window is related to the mean global surface temper- (case 2) given by Schwartzman (1999). The corresponding ature. If the global surface temperature is inside this win- parameters for the biospheres are summarized in Table 1. All dow a global abundance of biosphere type i is possible. Our other parameters have been taken from Franck et al. (2002) parameterization of the biological productivity is based on for the favoured model with spreading dependent hydrother- oxygenic photosynthesis. In this study we investigated two mal flux and constant pH of the ocean. The biotic enhance- different cases: first we applied the more restrictive tolerance ment factor β3 has been adjusted in such a way that complex windows given by von Bloh et al. (2003) and second the val- multicellular life appears spontaneously first at −542 Myr. ues given by Schwartzman (1999). He defines physiologi- We derived β3=3.6 in case 1 and β3=17.5 in case 2. cal tolerances for local temperatures of different organisms, In Fig. 2a we have plotted the results of case 1 for the evo- which are 15◦C to 20◦C higher than ours. The explicit values lution of the mean global surface temperature from the Ar- are given in Table 1. chaean to the long-term future in about 2 Gyr. Estimations

Biogeosciences, 3, 85–92, 2006 www.biogeosciences.net/3/85/2006/ S. Franck et al.: Causes and timing of future biosphere extinctions 89 of Precambrian palaeotemperatures date back to the early Archaean and are based on oxygen isotopic composition of cherts (Knauth and Lowe, 2003). According to these data, the ocean surface water has cooled from 70◦C(±15◦C) in the Archaean ocean to the present value. Such values are conceivable as mean global surface temperatures during the early Archaean when atmospheric CO2 levels could have been very high (Franck et al., 2002). Nevertheless, high Ar- chaean temperatures from the oxygen isotopic composition of cherts are still controversial (Sleep and Hessler, 2006). In our model high Archaean atmospheric CO2 levels are caused by two effects: first there is only a small amount of continental area for weathering (reduced sink of atmospheric CO2) and second there is an intense volcanic outgassing due to higher geodynamic activity (elevated source of atmospheric CO2). In Fig. 2b we show the corresponding cumulative biosphere pools. The question of how much biomass exists at different stages in the Earth’s evolution is of great importance for our modelling. The problem of the quantitative evolution of the terrestrial biomass through time is a question of scien- tific and practical concern, because fossil organic carbon is the prime energy source of the present society (Schidlowski, 1991). During the entire existence of the biosphere procaryotes are always present. 2 Gyr ago eucaryotic life first appears because the global surface temperature reaches the tolerance window for eucaryotes. This moment correlates with the onset of a rapid temperature drop caused by increasing continental area. The resulting increase in the weathering Fig. 3. Case 2, (a) Evolution of global surface temperature (solid flux takes out CO2 from the atmosphere. In contrast to the green line). The coloured area indicates the evolution of the nor- eucaryotes the first appearance of complex multicellular life malized continental area according to Condie (1990). (b) Evolution starts with an explosive increase in biomass connected with of the cumulative biosphere pools for procaryotes (red), eucaryotes a strong decrease in Cambrian global surface temperature (green), and complex multicellular life (brown). at about 0.54 Gyr ago. The biological colonization of land surface by metaphyta and the consequent increase in silicate weathering rates caused a reduction in atmospheric CO2 and contrast to the first appearance of complex multicellular life planetary cooling. After the Cambrian explosion there is a via the Cambrian explosion, its extinction proceeds more or continuous decrease of biomass in all pools. At 0.35 Gyr ago less continuously. there is a slight drop in all biomass pools connected with the In Fig. 3 we have plotted the results of case 2 for the evo- rise of vascular plants. The continuous decrease in biomass lution of the mean global surface temperature and the cumu- of primitive life forms (procaryotes and eucaryotes) since the lative biosphere pools from the Archaean to the long-term Cambrian explosion is related to the fact that Phanerozoic future in about 2 Gyr. There are no qualitative differences surface temperatures are below the optimum for these life to case 1. In particular, the Cambrian explosion event is even forms. The decrease in biomass of complex life forms is due more pronounced and the three biosphere types cease to exist to the fact that there is a continuous decrease in Phanero- in the same way. Complex multicellular life becomes extinct zoic atmospheric carbon content. At present the biomass is in about 1.2 Gyr and eucaryotes in about 1.5 Gyr. almost equally distributed between the three pools and the In both cases the ultimate life span of the biosphere, i.e. the mean global surface temperature of about 15◦C is near the extinction of procaryotes, ends at about 1.6 Gyr. In this case optimum value for complex multicellular life. the extinction is not caused by the temperature leaving the In the future we can observe a further continuous decrease tolerance window but by a too low atmospheric CO2 content of biomass with the strongest decrease in the complex multi- for photosynthesis. cellular life. The life spans of complex multicellular life and In Fig. 4 we have plotted for case 1 the time when the dif- of eucaryotes end at about 0.8 Gyr and 1.3 Gyr from present, ferent life forms appear and disappear and the time interval respectively. In both cases the extinction is caused by reach- in which perturbations may trigger the first emergence and ing the upper limit of the temperature tolerance window. In the extinction of complex life prematurely as a function of www.biogeosciences.net/3/85/2006/ Biogeosciences, 3, 85–92, 2006 90 S. Franck et al.: Causes and timing of future biosphere extinctions

Fig. 4. Stability diagram for the three types of biosphere (case 1) as a function of the biotic enhancement factor, β3. In the red area only procaryotic life exists while in the green area eucaryotic and procaryotic life coexist. In the brown area complex multicellular life appears additionally. The dashed area indicates the time interval in which a perturbation may trigger the first emergence or extinction of complex multicellular life prematurely. The horizontal dashed line denotes the time of the Cambrian explosion. The vertical dashed line denotes the corresponding value of β3. the biotic enhancement factor β3. Up to −1.75 Gyr there is To analyze the influence of the upper temperature tol- only a unique solution (no bistability). Therefore, the Huro- erance for complex multicellular life (Tmax,3) on the life nian glaciations circa 2.4 Gyr ago cannot trigger a prema- span in more detail we performed additional simulations for ◦ ◦ ◦ turely emergence of eucaryotic or even complex life. On 18.6 C≤Tmax,3≤55 C for case 2. For Tmax,3<18.6 C no the other hand, Neoproterozoic snowball Earth events have Cambrian explosion could appear. In all other cases the the potential to initiate an earlier appearance of complex life biotic enhancement factor β3 is adjusted to reproduce the forms. However, these global glaciations are followed by Cambrian explosion at the right time 542 Myr ago. For ◦ a global hothouse (Hoffman and Schrag, 2002) that imme- Tmax,3>30 C the ultimate life span stays almost constant at diately pushes the temperatures again above the upper tol- 1.6 Gyr from present. This is shown in Fig. 5. An investiga- erance limit. In the case of β3=3.6 complex multicellular tion of the influence of pmin on the ultimate life span resulted life could appear in principle at 1.7 Gyr ago. There is evi- an extension of only 100 Myr for a minimum CO2 pressure dence for small metazoans and multicellular algae well be- for photosynthesis of 1 ppm in case 1, while in case 2 the life fore the Ediacarans (appeared 600 Myr ago), but they had span was extended only by 20 Myr. no influence on weathering and possibly a higher tempera- The diverse causes of the future biosphere extinction can ture tolerance. For β3<3.6 complex multicellular life had to also be derived from the so-called “terrestrial life corridor” appear first before the Cambrian era. For β3>3.6 a pertur- (TLCi) for the different life forms: bation in environmental conditions is necessary to force the TLC := p ,T 5 p ,T > 0 . (15) appearance of complex multicellular life in the Cambrian. i CO2 s i CO2 s For β3>16 eucaryotes and complex multicellular life would In Fig. 6 we show the atmospheric carbon dioxide content appear simultaneously. Another important result is that for (black line) over time from the Archaean up to the long- β3>6.38 complex multicellular life cannot appear sponta- term future for the three types of biosphere for case 1. In neously but only due to cooling events, because the Earth the non-coloured region of Fig. 5 no biosphere may exist surface temperature always remains above the upper temper- because of inappropriate temperature or atmospheric carbon ◦ ature tolerance of 30 C for complex multicellular life. dioxide content. The coloured domain is the cumulative In contrast to the Neoproterozoic, in the future there will TLC for the three biosphere pools in analogy to Fig. 2b. be no bistability for values β3<5, i.e. the extinction of com- Again we can see that complex multicellular life and eu- plex multicellular life will not proceed as an implosion (in caryotes become extinct in about 0.8 Gyr and 1.3 Gyr, re- comparison to the Cambrian explosion). Our results refine spectively, because of inappropriate temperature conditions. the predictions of Ward and Brownlee (2002). The procaryotes become extinct in about 1.6 Gyr because

Biogeosciences, 3, 85–92, 2006 www.biogeosciences.net/3/85/2006/ S. Franck et al.: Causes and timing of future biosphere extinctions 91

Fig. 5. Life spans of procaryotic, eucaryotic, and complex multicellular life as a function of the upper temperature tolerance of complex multicellular life, Tmax,3, for case 2. The bold dashed line denotes the necessary biotic enhancement factor, β3, to adjust the Fig. 6. The evolution of atmospheric CO2 concentration in units of ◦ Cambrian explosion time. Note, that for Tmax,3<18.6 C (vertical present atmospheric level (PAL) (black line) for case 1. The brown dotted line) no Cambrian explosion can appear. + green + red coloured region defines the terrestrial life corridor (TLC) for procaryotes. The green + brown coloured region defines the TLC for procaryotes and eucaryotes in coexistence. The brown coloured region is the TLC where all three biosphere types may of achieving the minimum value for atmospheric CO2 content. Our estimation is valid only for photosynthesis-based exist together. life. Other life forms like chemolithoautotrophic hyperther- mophiles may extend the ultimate life span. Acknowledgements. We want to thank T. Lenton and an anony- mous reviewer for the constructive remarks. This work was 4 Conclusions supported by the German Science Foundation (DFG, grant number Fr 910/10-3), and the University Science Program (HWP, grant Procaryotes, eucaryotes, and complex multicellular life number 24#2597-04/333). forms will become extinct in reverse sequence of their appearance. This is a quantitative manifestation of the quali- Edited by: F. Westall tative predictions of Ward and Brownlee (2002). We have shown that nonlinear interactions in the biosphere-geosphere References system cause bistability during the Neo- and Mesoprotero- zoic era. For realistic values of the biotic enhancement of Bounama, C., Franck, S., and von Bloh, W.: The fate of Earth’s weathering there is no bistability in the future solutions for ocean, Hydrol. Earth Syst. Sci., 5, 569–575, 2001. complex life. Therefore, complex organisms will not become Caldeira, K. and Kasting, J. F.: The life span of the biosphere revis- extinct by an implosion (in comparison to the Cambrian ex- ited, Nature, 360, 721–723, 1992. plosion). Eucaryotes and complex life become extinct be- Condie, K. C.: Growth and accretion of continental crust: infer- cause of too high surface temperatures in the future. The ences based on Laurentia, Chem. Geol., 83, 183–194, 1990. time of extinction is mainly determined by the upper tem- Covindasamy, B. and Caldeira, K.: Geoengeneering Earth’s radiation balance to mitigate CO -induced climate change, Geophys. perature tolerance limit of these life forms. The ultimate life 2 Res. Lett., 27, 2141, 2000. span of the biosphere is defined by the extinction of procary- Franck, S. and Bounama, C.: Effects of water-dependent creep rate otes in about 1.6 Gyr because of CO2 starvation. Only in a on the volatile exchange between mantle and surface reservoirs, small fraction (1.3–1.7 Gyr) of its habitability time (6.2 Gyr) Phys. Earth Planet. Inter., 92, 57–65, 1995. can our home planet harbour advanced life forms. Franck, S., Block, A., von Bloh, W., Bounama, C., Schellnhuber, However, humankind could extend its life time on Earth H.-J., and Svirezhev, Y.: Reduction of biosphere life span as a by geoengineering (Hoffert et al., 2002). For example, a consequence of geodynamics, Tellus, 52B, 94–107, 2000. large mirror at the Lagrange point L1 of the Sun-Earth sys- Franck, S., Kossacki, K. J., von Bloh, W., and Bounama, C.: Long- tem could deflect partly the solar flux and stabilize the future term evolution of the global carbon cycle: Historic minimum climate (Covindasamy and Caldeira, 2000). of global surface temperature at present, Tellus, 54B, 225–343, 2002. Hoffert, M. I., Caldeira, K., Benford, G., Criswell, D. R., Green, C., Herzog, H., Jain, A. K., Kheshgi, H. S., Lackner, K. S., Lewis, J. S., Lightfoot, H. D., Manheimer, W., Mankins, J. C., Mauel, M. www.biogeosciences.net/3/85/2006/ Biogeosciences, 3, 85–92, 2006 92 S. Franck et al.: Causes and timing of future biosphere extinctions

E., Perkins, L. J., Schlesinger, M. E., Volk, T., and Wigley, T. M. Lovelock, J. E. and Whitﬁeld, M.: The life span of the biosphere, L.: Advanced technology path to global climate stability: energy Nature, 296, 561–563, 1982. for a greenhouse planet, Science, 298, 981–987, 2002. Schidlowski, M.: Quantitative evolution of global biomass through Hoffman, P. F. and Schrag, D. P.: The snowball Earth hypothesis: time: Biological and geochemical constraints, in: Scientists on testing the limits of global change, Terra Nova, 14(3), 129–155, Gaia, edited by: Schneider, S. H. and Boston, P. J., MIT Press, 2002. Cambridge, 211–222, 1991. Kasting, J. F., Whitmire, D. P., and Reynolds, R. T.: Habitable zones Schwartzman, D. W.: Life, temperature and the Earth: The self- around main sequence stars, Icarus, 101, 108–128, 1993. organizing biosphere, Columbia University Press, New York, Knauth, L. P. and Lowe, D. R.: High Archean climatic temperature 1999. inferred from oxygen isotope geochemistry of cherts in the 3.5 Sleep, N. H. and Hessler, A. M.: Weathering of quartz as an Ga Swaziland Supergroup, South Afrika, GSA Bulletin, 115(5), Archean climatic indicator, Earth Planet. Sci. Lett., 241, 594– 566–580, 2003. 602, 2006. Lenton, T. M.: Land and ocean carbon cycle feedback effects on Unsold,¨ A.: Der Neue Kosmos, Springer, Berlin, 1967. global warming in a simple Earth system model, Tellus, 52B, Von Bloh, W., Bounama, C., and Franck, S.: Cambrian explo- 1159–1188, 2000. sion triggered by geosphere-biosphere feedbacks, Geophys. Res. Lenton, T. M. and von Bloh, W.: Biotic feedback extends the life Lett., 30(18), 1963, doi:10.1029/2003GL017928, 2003. span of the biosphere, Geophys. Res. Lett., 28(9), 1715–1718, Ward, P. and Brownlee, D.: The life and death of planet Earth, Pi- 2001. atkus, London, 2002.

Biogeosciences, 3, 85–92, 2006 www.biogeosciences.net/3/85/2006/